M45, aka the Pleiades

[xilman]

Tom gave me nine images of the Pleiades. Six are of adequate quality though one is a short exposure and so very noisy. One has every star image doubled, apparently because the camera mount was knocked part way through the exposure. The remaining pair are trailed.

I'm now part way through processing these images. I've deconvolved one of the trailed images with Lucy-Richardson and using a position dependent PSF. This was then stacked with two of the good images. The R, G and B channels were processed independently. The result is attached.

I'm very impressed that a standard DSLR show the reflection nebulosity very clearly, even though the images were taken from a heavily light polluted site and, further, that this is still a preliminary result.

Still to do is to deconvolve the remaining trailed image and to add it and the remaining good images to the stack. After that, further image restoration (probably using the maximum entropy method to retain the low-contrast nebulosity while sharpening the stars), followed first by colour balancing to turn the sky black and then background removal to remove both vignetting and residual light pollution.

Given the quality of the intermediate image, I've high hopes for the final result.

[fivemack]

Excellent work! I'm very impressed how much you have managed to pull out of the browny-orange murk (which gets to 2% of saturation per second of exposure at f/3.5 ISO 2000, and those Pleiades shots are at f/2.8 where the effect is worse) that is the north-west Cambridge sky.

[xilman]

Originally Posted by fivemack

Excellent work! I'm very impressed how much you have managed to pull out of the browny-orange murk (which gets to 2% of saturation per second of exposure at f/3.5 ISO 2000, and those Pleiades shots are at f/2.8 where the effect is worse) that is the north-west Cambridge sky.

Which only goes to show that avoiding saturation is a good thing. Sky glow is largely uncorrelated noise and can be reduced by a factor of sqrt(#images) without any fancy processing as long as the pixels are not saturated. I have been assuming for photometry purposes that your detector has a linear response when the intensity lies between 300 and 40K. That appears to produce mostly acceptable PSF estimates.

Lots of short exposures are better than a few over-saturated images for making pretty pictures. I don't yet have enough experience to say whether that is also true for accurate photometry, though I do know that saturated star images can not be accurately measured.

[xilman]

Here's the latest result. All the images co-added weighted by exposure time. Not properly colour balanced yet (the stars are Na sky glow colour in the raw image) but munged to emphasize the reflection nebulae. The munging turns everything blue. A logarithmic intensity scale is chosen and there's some gentle contrast stretching.

For comparison is the same region from the DSS. I'm quite pleased how some of the structure in the nebulae has appeared, especially around Merope, and the dark cross in the centre of the main parallelogram.

[VictordeHolland]

Following this thread .

I'm surprised by the amount of detail, like part of the diffusion nebulae. Did you attach a telescope to a full-frame DSLR camera? Or is this a (normal/consumer) camera on a tri-pod?

If you could somehow make the sky black and get rid of some noise, while still retaining the nebulae, it would look even nicer. It can't compete with pics from Hubble or sky surveys, but hé, they're using multi-million dollar equipment.....

[fivemack]

Originally Posted by VictordeHolland

Following this thread .

I'm surprised by the amount of detail, like part of the diffusion nebulae. Did you attach a telescope to a full-frame DSLR camera? Or is this a (normal/consumer) camera on a tri-pod?

If you could somehow make the sky black and get rid of some noise, while still retaining the nebulae, it would look even nicer. It can't compete with pics from Hubble or sky surveys, but hé, they're using multi-million dollar equipment.....

Those images were from a 200mm-focal-length f/2.8 lens on a Nikon D90, on a traditional tripod and a low-end tracking mount. The gigabytes of images I gave to Xilman include some using a 400mm-focal-length f/3.5 lens (which is 2.8kg in weight and the kind of thing you usually see pointed at lions on safari - https://www.mir.com.my/rb/photograph...tos/400mm1.htm), a D750 with a larger and more sensitive sensor, and a significantly better tripod and mount.

I have a modest telescope (Celestron 6, 1500mm focal length) and have taken a few photos through it, but have had real trouble getting the mount to point at things accurately enough to get them into the telescope's field of view; and working on my own in the cold in November I'm not very keen on switching between the telescope and a conventional lens repeatedly until I'd got the pointing sorted out. So those pictures amount to a few exposures of half the Pleiades and a few of a chunk of Orion's belt. It's a much longer focal ratio (f/10) so more suited to stars than to nebulosity.

[xilman]

Originally Posted by VictordeHolland

If you could somehow make the sky black and get rid of some noise, while still retaining the nebulae, it would look even nicer.

That's next on my to-do list. Crude "photoshopping" will do what you ask but I want to preserve as much of the original data as possible --- no cheating in other words.

All of this is an educational exercise for me. I want to learn what information can be extracted from the data, even though the data is inevitably going to be poorer than that taken with professional equipment from a first class site. Frankly, I'm astonished that the Merope nebula shows structure consistent with that found in images taken in much less severe conditions. I would have been pleased just to see the nebula at all. Perhaps Tom or I should post one of the raw images to show what I've had to work with.

Added in edit: attached one of the better raw images. Note the Na light pollution.

[xilman]

Originally Posted by xilman

That's next on my to-do list.

The very first attempt produced something quite beautiful! So pretty that I'll save it for posterity, not that posterity has ever done anything for me.

The background is a grainy orange-brown but overlaid is an ethereal pale blue nebulosity, exactly where it should be, and showing much more detail than before.

[xilman]

Originally Posted by xilman

The very first attempt produced something quite beautiful! So pretty that I'll save it for posterity, not that posterity has ever done anything for me.

Rather than clog up Mike's server with yet more data, here is the image on my Twitter post.

[Xyzzy]

Originally Posted by xilman

Rather than clog up Mike's server with yet more data, here is the image on my Twitter post.

Feel free to upload them here. We have plenty of room!

[xilman]

Originally Posted by Xyzzy

Feel free to upload them here. We have plenty of room!

Ok, here you are. The original FITS has been converted first to PNG and then JPG, so some quality has been lost. Nonetheless, I'm still amazed at the quality of the image derived from Tom's data.